CN112488499A - Judicial risk early warning method, device and terminal based on supplier scale - Google Patents
Judicial risk early warning method, device and terminal based on supplier scale Download PDFInfo
- Publication number
- CN112488499A CN112488499A CN202011361506.7A CN202011361506A CN112488499A CN 112488499 A CN112488499 A CN 112488499A CN 202011361506 A CN202011361506 A CN 202011361506A CN 112488499 A CN112488499 A CN 112488499A
- Authority
- CN
- China
- Prior art keywords
- information
- judicial
- risk
- supplier
- scale
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000008569 process Effects 0.000 claims abstract description 23
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 238000013136 deep learning model Methods 0.000 claims abstract description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 47
- 238000012549 training Methods 0.000 claims description 41
- 238000012360 testing method Methods 0.000 claims description 19
- 238000011176 pooling Methods 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 9
- 238000013058 risk prediction model Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 3
- 230000002087 whitening effect Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 2
- 210000002569 neuron Anatomy 0.000 description 9
- 230000006870 function Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000003213 activating effect Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 230000003247 decreasing effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0609—Buyer or seller confidence or verification
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Development Economics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the field of bid and tender management, and provides a supplier scale-based judicial risk early warning method, a supplier scale-based judicial risk early warning device and a supplier scale-based judicial risk early warning terminal, which are used for solving the problem of change of the judicial risk of a supplier in the bid and tender process. The judicial risk early warning method based on the scale of the suppliers, provided by the invention, comprises the following steps: acquiring bidding purchasing process information; acquiring scale information and judicial information of a supplier in a key link in a tendering and purchasing process, wherein the judicial information comprises one or more of information of a distrusted executor, information of a court announcement, information of a referee document and information of an executive announcement; preprocessing the scale information of the suppliers and the judicial information to obtain input information, and inputting the input information into a judicial risk model, wherein the judicial risk model is a deep learning model; an output of a judicial risk level for the provider, the judicial risk level comprising one of a low risk and a high risk. And loss caused by excessive judicial risk in the bidding process is avoided or reduced.
Description
Technical Field
The invention relates to the field of bid and tender management, in particular to a judicial risk early warning method based on the scale of a supplier.
Background
According to the overall requirements of 'notice on the analysis table of advanced bidding management reform tasks of the issuing company' of the file No. 2019 of the radio and television enterprise '8', intelligent recommendation, risk analysis and intelligent early warning are realized by utilizing the technologies of supplier data reconstruction and the like, the high compliance efficiency of the selected suppliers for bidding purchase is ensured, and the risks of performing and auditing caused by the self risks of the suppliers in the purchasing process are prevented.
Risks in the bidding process are now mainly obtained by a skilled person, such as a lawyer, making a due diligence on the background of the supplier. The employment of professionals such as a lawyer and the like to conduct investigation consumes a large amount of manpower and material resources, and has extremely high cost. Some existing third-party enterprise survey companies can evaluate risks of enterprises, but the evaluations are not comprehensive or have low speciality, provided services can only be performed on webpages, evaluation capacity is limited, and adjustment cannot be performed according to needs. At present, computer software capable of evaluating the judicial risk of a supplier in the bidding purchasing process is lacked, and the real-time judicial risk information of the supplier cannot be known in time in the bidding purchasing process, particularly in key links such as project evaluation, scaling and the like.
Disclosure of Invention
The technical problem solved by the invention is the problem of change of judicial risk of the supplier in the bidding process, and the judicial risk early warning method based on the scale of the supplier is provided.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
a judicial risk early warning method based on supplier scale comprises the following steps:
acquiring bidding purchasing process information;
acquiring scale information and judicial information of a supplier in a key link in a tendering and purchasing process, wherein the judicial information comprises one or more of information of a distrusted executor, information of a court announcement, information of a referee document and information of an executive announcement;
preprocessing the scale information of the suppliers and the judicial information to obtain input information, and inputting the input information into a judicial risk model, wherein the judicial risk model is a deep learning model;
an output of a judicial risk level for the provider, the judicial risk level comprising one of a low risk and a high risk.
And judging the judicial risk of the supplier according to the key information of the process of tendering purchase, and providing reference opinions for tendering purchase at a key time node in time.
The loss can be stopped in time, and the loss caused by excessive judicial risk is avoided or reduced.
Preferably, the scale information includes one or more of total supplier asset information, total supplier revenue information, and total supplier profit information. The inventors have found that using scale information as input can greatly improve the accuracy of risk level determination.
Preferably, the judicial risk model is a convolutional network model, and the training method of the convolutional network model is as follows:
acquiring training sample data, inputting the sample data into a convolutional neural network model for training to obtain an initial risk prediction model;
and acquiring test sample data, and inputting the test sample data into the initial risk prediction model for evaluation test. The accuracy rate of judging the enterprise judicial risk by the trained convolutional neural network model is high.
Preferably, the training sample data and the test sample data are part of sample information of a historical provider, and the sample information is scale information, judicial information and judicial risk level of the historical provider; the training sample data is: the test sample data = 2-3: 1
And (4) performing picture processing on the sample information, cutting the sample information into uniform size, and performing whitening processing on the cut sample.
Preferably, the convolutional neural network comprises a first convolutional neural network and a second convolutional neural network;
inputting provider scale information in training sample data into a first convolutional neural network for training, wherein the first convolutional neural network comprises a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected;
adding a third full-link layer and a fourth full-link layer to the trained first convolutional neural network to generate a second convolutional neural network, wherein the third full-link layer is the same as the trained first full-link layer and is connected with the second pooling layer, and the fourth full-link layer is the same as the trained second full-link layer and is connected with the third full-link layer;
and training the second convolutional neural network according to the training sample data. Two connected neural networks are arranged, so that the accuracy of risk judgment can be enhanced.
Preferably, the method of training the second convolutional neural network comprises:
and training the second convolutional neural network by taking the supplier scale information and the judicial information in the training sample as the input of the first convolutional layer in the second convolutional network and taking the judicial risk level as the output of the second convolutional neural network. And the input and the output of the second neural network are adjusted, so that the accuracy of risk judgment can be further improved.
Preferably, the convolutional neural network is optimized according to the result of the evaluation test, and the optimization method comprises the following steps:
and (4) increasing or reducing the number of convolution layers on the structure of the convolution neural network, and retraining. Different bidding items, different suppliers and different risks may occur, and the number of convolutional layers may be increased or decreased as necessary to improve the accuracy of the determination.
Judicial risk early warning device based on supplier scale includes:
the progress information acquisition module acquires bidding purchase information;
the system comprises an information acquisition module, a data processing module and a data processing module, wherein the information acquisition module acquires scale information and judicial information of a supplier in a key link of bidding purchase, and the judicial information comprises one or more of information of a message loss executor, court announcement information, official document information and execution announcement information;
the preprocessing module is used for preprocessing the supplier scale information and the judicial information to obtain input information;
the risk judgment module inputs input information into a judicial risk model, and the judicial risk model is a deep learning model;
an output module that outputs the supplier's judicial risk level.
The judicial risk early warning terminal based on the scale of suppliers comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor is used for executing the computer program to execute the method.
A supplier-scale judicial risk early warning storage medium storing a computer program executable to perform the method as described above.
Compared with the prior art, the invention has the beneficial effects that: the loss can be stopped in time, and the loss caused by excessive judicial risk in the bidding process is avoided or reduced.
Drawings
Fig. 1 is a schematic diagram of a judicial risk early warning method based on vendor scale.
FIG. 2 is a schematic diagram of a first convolutional neural network.
Fig. 3 is a schematic diagram of a second convolutional neural network.
Detailed Description
The following examples are further illustrative of the present invention and are not intended to be limiting thereof.
Example 1
A forensic risk early warning method based on vendor scale, in some embodiments of the present application, comprises:
acquiring bidding purchasing process information;
acquiring scale information and judicial information of a supplier in a key link in a tendering and purchasing process, wherein the judicial information comprises one or more of information of a distrusted executor, information of a court announcement, information of a referee document and information of an executive announcement;
preprocessing the scale information of the suppliers and the judicial information to obtain input information, and inputting the input information into a judicial risk model, wherein the judicial risk model is a deep learning model;
an output of a judicial risk level for the provider, the judicial risk level comprising one of a low risk and a high risk.
And judging the judicial risk of the supplier according to the key information of the process of tendering purchase, and providing reference opinions for tendering purchase at a key time node in time.
The loss can be stopped in time, and the loss caused by excessive judicial risk is avoided or reduced.
In some embodiments of the present application, after the solicitation purchase process information is obtained; and acquiring a key link in the bid procurement process, wherein in the embodiment, a project evaluation link and a calibration link are taken as key links.
It should be understood that bidding projects vary, bidding suppliers vary, and the corresponding bidding process has different key element settings.
In some embodiments of the present application, the scale information includes one or more of total supplier asset information, total supplier revenue information, and total supplier profit information.
The inventors have found that using scale information as input can greatly improve the accuracy of risk level determination.
In some embodiments of the present application, the judicial risk model is a convolutional network model, and the training method of the convolutional network model is:
acquiring training sample data, inputting the sample data into a convolutional neural network model for training to obtain an initial risk prediction model;
and acquiring test sample data, and inputting the test sample data into the initial risk prediction model for evaluation test.
The accuracy rate of judging the enterprise judicial risk by the trained convolutional neural network model is high.
In some embodiments of the present application, the training sample data and the test sample data are part of sample information of a historical provider, the sample information being scale information, judicial information and judicial risk level of the historical provider; the training sample data is: the test sample data = 2-3: 1;
and (4) performing picture processing on the sample information, cutting the sample information into uniform size, and performing whitening processing on the cut sample.
In some embodiments of the present application, a database of vendor risk sample data is created, and risk elements are determined, in this example, the vendor risk elements include: whether the information is classified as information of a message loss executor, whether there is court announcement information, whether there is official document information, whether there is execution announcement information, whether the total asset of the supplier reaches a threshold value, whether the total income of the supplier suddenly decreases, and whether the total profit of the supplier suddenly decreases.
In some embodiments of the present application, the judicial risk classification is low risk and high risk.
In some embodiments of the present application, the training sample data: the test sample data =2: 1.
In some embodiments of the present application, sample information from a history provider is produced as a picture, the picture is uniformly cropped to 100px × 100px, a central region is cropped for evaluation or random cropping for training, and the picture is whitened.
In some embodiments of the present application, the convolutional neural network comprises a first convolutional neural network and a second convolutional neural network;
inputting provider scale information in training sample data into a first convolutional neural network for training, wherein the first convolutional neural network comprises a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected;
adding a third full-link layer and a fourth full-link layer to the trained first convolutional neural network to generate a second convolutional neural network, wherein the third full-link layer is the same as the trained first full-link layer and is connected with the second pooling layer, and the fourth full-link layer is the same as the trained second full-link layer and is connected with the third full-link layer;
and training the second convolutional neural network according to the training sample data.
Two connected neural networks are arranged, so that the accuracy of risk judgment can be enhanced. And the input and the output of the second neural network are adjusted, so that the accuracy of risk judgment can be further improved.
In some embodiments of the present application, a sample picture of a supplier for training is cropped to a size of 100px × 100px, and the supplier's judicial risk rating is low risk. The first convolution layer has 20 convolution kernels, the number of parameters of each convolution kernel is 3 × 6 × 6, and 3 convolution kernels with the size of 6 × 6 are convolved in each channel, and the step size is 1. After the convolution of the first convolution layer, it can be known from (100-6)/1 +1=95 that the size of the obtained image is 95px × 95px, that is, 20 feature maps with a size of 95px × 95px are obtained, and actually, after the convolution of the first convolution layer, it is necessary to adjust the output of the first convolution layer by activating a function, so as to avoid that the output of the next layer is a linear combination of the previous layer and cannot approach any function. The overfitting problem is further alleviated by using a relu (rectified Linear unit) function as the activation function, which is expressed by f (x) max (0, w · x + b), where w · x + b is a conventional Linear function. And then entering a first pooling layer, and sub-sampling the image by utilizing the principle of local image correlation, thereby reducing data processing and retaining useful information. Here, pooling is performed by maximum overlap pooling, i.e., 95px × 95px feature map is partitioned into blocks, each block has a size of 3 × 3 and a step size of 2, and the maximum value of each block is counted as a pixel value of the pooled image. From (95-3)/2+1 ═ 47, it can be seen that the feature map size after pooling is 47px × 47px, and after the first downsampling layer, 20 feature maps of 47px × 47px were obtained.
The 20 feature maps of 47px enter the second convolutional layer, which has 48 convolution kernels, the number of parameters of each convolution kernel is 6 × 6, and 1 convolution kernel with the size of 6 × 6 is convolved in each channel, and the step size is 1. Then, after the second convolutional layer convolution, the size of the obtained image is 42px × 42px as shown by (47-6)/1 +1= 42. In the convolution process of the second convolution layer, 20 pieces of 47px × 47px are weighted and combined, and finally 48 pieces of 42px × 42px feature maps are obtained. And activating the output of the second convolution layer through a ReLU function, and entering a second down-sampling layer. According to the maximum overlapping pooling principle, the 42px × 42px feature map is partitioned into blocks, the size of each block is 2 × 2, the step size is 2, and the maximum value of each block is counted as the pixel value of the pooled image. From (42-2)/2+1 ═ 21, it can be seen that the feature map size after pooling is 21px × 21px, and 48 feature maps of 21px × 21px were obtained after the second downsampling layer.
And then entering a first full connection layer, wherein 512 neurons of the first full connection layer are selected, and the output of the first full connection layer is 512 characteristic graphs with the size of 1px multiplied by 1 px. In actual processing, the 512 feature maps are usually activated by the ReLU function, and then subjected to dropout processing, where dropout is understood as model averaging, that is, when conducting forward in the training process, the activation value of a certain neuron stops working with a certain probability p, that is, the activation value of the neuron becomes 0 with the probability p. In this embodiment, the number of neurons in the first fully-connected layer is 512, and the dropout ratio is selected to be 0.5, so that after the layer of neurons passes through dropout, the value of about 256 neurons therein is set to 0, and the overfitting is relieved by preventing the synergistic effect of certain features, so as to avoid the phenomenon that one neuron appears to be connected with another neuron. And inputting the characteristic diagram after dropout processing into a second full connection layer, wherein the number of the neurons of the second full connection layer is 2, and finally the output of the second full connection layer is also 2, which respectively correspond to the high risk probability and the low risk probability. And adjusting the output of the second full-connection layer according to the judicial risk level corresponding to the input sample image as the predicted result, and performing back propagation according to a method of minimizing errors to adjust each parameter in the first convolutional neural network. After a large amount of sample image data are trained, a trained first convolution neural network is obtained.
And adding a third full-link layer and a fourth full-link layer in the trained first convolutional neural network to generate a second convolutional neural network, wherein the third full-link layer is the same as the trained first full-link layer and is connected with the second pooling layer, and the fourth full-link layer is the same as the trained second full-link layer and is connected with the third full-link layer. A first branch with the second full connection layer as output and a second branch with the fourth full connection layer as output are formed
In some embodiments of the present application, a method of training a second convolutional neural network comprises:
and training the second convolutional neural network by taking the supplier scale information and the judicial information in the training sample as the input of the first convolutional layer in the second convolutional network and taking the judicial risk level as the output of the second convolutional neural network.
In some embodiments of the present application, the second convolutional neural network is trained based on the sample image data. And reclassifying the risk categories to obtain a first risk category and a second risk category, wherein the first risk category is one of low risk and other risk, and the second risk category is one of high risk and other risk. And taking the sample image as the number of first convolution layers of the second convolution network, taking the first risk category as the output of a second full-connection layer in the second convolution neural network, taking the second risk category as the output of a fourth full-connection layer in the second convolution neural network, and training the second convolution neural network.
In some embodiments of the present application, the convolutional neural network is optimized according to the result of the evaluation test, and the optimization method includes:
and (4) increasing or reducing the number of convolution layers on the structure of the convolution neural network, and retraining.
Different bidding items, different suppliers and different risks may occur, and the number of convolutional layers may be increased or decreased as necessary to improve the accuracy of the determination.
The judicial risk early warning device based on the scale of the supplier comprises the following components in some embodiments of the application:
the progress information acquisition module acquires bidding purchase information;
the system comprises an information acquisition module, a data processing module and a data processing module, wherein the information acquisition module acquires scale information and judicial information of a supplier in a key link of bidding purchase, and the judicial information comprises one or more of information of a message loss executor, court announcement information, official document information and execution announcement information;
the preprocessing module is used for preprocessing the supplier scale information and the judicial information to obtain input information;
the risk judgment module inputs input information into a judicial risk model, and the judicial risk model is a deep learning model;
an output module that outputs the supplier's judicial risk level.
A supplier-scale-based judicial risk early warning terminal, in some embodiments of the present application, comprises a processor and a memory, the memory having stored therein a computer program, the processor being configured to execute the computer program to perform the method described above.
In some embodiments of the present application, a computer program is stored that can be executed to implement the above-described method.
The above detailed description is specific to possible embodiments of the present invention, and the above embodiments are not intended to limit the scope of the present invention, and all equivalent implementations or modifications that do not depart from the scope of the present invention should be included in the present claims.
Claims (10)
1. A judicial risk early warning method based on supplier scale is characterized by comprising the following steps:
acquiring bidding purchasing process information;
acquiring scale information and judicial information of a supplier in a key link in a tendering and purchasing process, wherein the judicial information comprises one or more of information of a distrusted executor, information of a court announcement, information of a referee document and information of an executive announcement;
preprocessing the scale information of the suppliers and the judicial information to obtain input information, and inputting the input information into a judicial risk model, wherein the judicial risk model is a deep learning model;
an output of a judicial risk level for the provider, the judicial risk level comprising one of a low risk and a high risk.
2. The supplier-scale-based judicial risk early warning method of claim 1, wherein the scale information comprises one or more of supplier total asset information, supplier total revenue information, and supplier total profit information.
3. The supplier scale-based judicial risk pre-warning method according to claim 2, wherein the judicial risk model is a convolutional network model, and the training method of the convolutional network model is as follows:
acquiring training sample data, inputting the sample data into a convolutional neural network model for training to obtain an initial risk prediction model;
and acquiring test sample data, and inputting the test sample data into the initial risk prediction model for evaluation test.
4. The supplier-scale-based judicial risk early warning method according to claim 1, wherein the training sample data and the test sample data are part of sample information of a historical supplier, and the sample information is scale information, judicial information and judicial risk level of the historical supplier; the training sample data is: the test sample data = 2-3: 1
And (4) performing picture processing on the sample information, cutting the sample information into uniform size, and performing whitening processing on the cut sample.
5. The vendor-scale-based judicial risk early warning method of claim 1, wherein the convolutional neural network comprises a first convolutional neural network and a second convolutional neural network;
inputting provider scale information in training sample data into a first convolutional neural network for training, wherein the first convolutional neural network comprises a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected;
adding a third full-link layer and a fourth full-link layer to the trained first convolutional neural network to generate a second convolutional neural network, wherein the third full-link layer is the same as the trained first full-link layer and is connected with the second pooling layer, and the fourth full-link layer is the same as the trained second full-link layer and is connected with the third full-link layer;
and training the second convolutional neural network according to the training sample data.
6. The vendor-scale-based judicial risk early warning method of claim 1, wherein the method of training the second convolutional neural network comprises:
and training the second convolutional neural network by taking the supplier scale information and the judicial information in the training sample as the input of the first convolutional layer in the second convolutional network and taking the judicial risk level as the output of the second convolutional neural network.
7. The supplier scale-based judicial risk early warning method according to claim 1, wherein the convolutional neural network is optimized according to the result of the evaluation test, the optimization method being:
and (4) increasing or reducing the number of convolution layers on the structure of the convolution neural network, and retraining.
8. Judicial risk early warning device based on supplier scale is characterized by comprising:
the progress information acquisition module acquires bidding purchase information;
the system comprises an information acquisition module, a data processing module and a data processing module, wherein the information acquisition module acquires scale information and judicial information of a supplier in a key link of bidding purchase, and the judicial information comprises one or more of information of a message loss executor, court announcement information, official document information and execution announcement information;
the preprocessing module is used for preprocessing the supplier scale information and the judicial information to obtain input information;
the risk judgment module inputs input information into a judicial risk model, and the judicial risk model is a deep learning model;
an output module that outputs the supplier's judicial risk level.
9. A terminal, comprising a processor and a memory, the memory having stored thereon a computer program, the processor being configured to execute the computer program to perform the method of any of claims 1 to 7.
10. A storage medium storing a computer program executable to perform the method of any one of claims 1 to 7 when executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011361506.7A CN112488499A (en) | 2020-11-28 | 2020-11-28 | Judicial risk early warning method, device and terminal based on supplier scale |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011361506.7A CN112488499A (en) | 2020-11-28 | 2020-11-28 | Judicial risk early warning method, device and terminal based on supplier scale |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112488499A true CN112488499A (en) | 2021-03-12 |
Family
ID=74936459
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011361506.7A Pending CN112488499A (en) | 2020-11-28 | 2020-11-28 | Judicial risk early warning method, device and terminal based on supplier scale |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112488499A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115133A1 (en) * | 2001-12-13 | 2003-06-19 | Dun & Bradstreet, Inc. | Higher risk score for identifying potential illegality in business-to-business relationships |
| CN106295521A (en) * | 2016-07-29 | 2017-01-04 | 厦门美图之家科技有限公司 | A kind of gender identification method based on multi output convolutional neural networks, device and the equipment of calculating |
| US20190266533A1 (en) * | 2018-02-26 | 2019-08-29 | Coupa Software Incorporated | Automated information retrieval based on supplier risk |
| CN111091271A (en) * | 2019-11-25 | 2020-05-01 | 上海欧冶采购信息科技有限责任公司 | Bidding risk early warning system and method |
| CN111292007A (en) * | 2020-02-28 | 2020-06-16 | 中国工商银行股份有限公司 | Supplier financial risk prediction method and device |
-
2020
- 2020-11-28 CN CN202011361506.7A patent/CN112488499A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115133A1 (en) * | 2001-12-13 | 2003-06-19 | Dun & Bradstreet, Inc. | Higher risk score for identifying potential illegality in business-to-business relationships |
| CN106295521A (en) * | 2016-07-29 | 2017-01-04 | 厦门美图之家科技有限公司 | A kind of gender identification method based on multi output convolutional neural networks, device and the equipment of calculating |
| US20190266533A1 (en) * | 2018-02-26 | 2019-08-29 | Coupa Software Incorporated | Automated information retrieval based on supplier risk |
| CN111091271A (en) * | 2019-11-25 | 2020-05-01 | 上海欧冶采购信息科技有限责任公司 | Bidding risk early warning system and method |
| CN111292007A (en) * | 2020-02-28 | 2020-06-16 | 中国工商银行股份有限公司 | Supplier financial risk prediction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108564286B (en) | Artificial intelligent financial wind-control credit assessment method and system based on big data credit investigation | |
| CN109859052B (en) | Intelligent recommendation method and device for investment strategy, storage medium and server | |
| CN112559900B (en) | Product recommendation method and device, computer equipment and storage medium | |
| Nyahokwe et al. | The impact of exchange rate volatility on South African exports | |
| CN112036631B (en) | Purchasing quantity determining method, purchasing quantity determining device, purchasing quantity determining equipment and storage medium | |
| CN110659926A (en) | Data value evaluation system and method | |
| CN113327164A (en) | Risk control method and device for futures trading and computer equipment | |
| CN107220246B (en) | Business object analysis method and device | |
| CN114782123A (en) | Credit assessment method and system | |
| Wang et al. | Fuzzy option pricing based on fuzzy number binary tree model | |
| CN116151840B (en) | User service data intelligent management system and method based on big data | |
| CN112488499A (en) | Judicial risk early warning method, device and terminal based on supplier scale | |
| Li et al. | Do bubbles alter contributions to price discovery? Evidence from the Chinese soybean futures and spot markets | |
| US20200311761A1 (en) | System and method for analyzing the effectiveness and influence of digital online content | |
| Jakob | Estimating correlation parameters in credit portfolio models under time-varying and nonhomogeneous default probabilities | |
| CN116150477A (en) | Financial information personalized recommendation method, device, equipment and medium | |
| CN115860505A (en) | Object evaluation method and device, terminal equipment and storage medium | |
| CN111178927A (en) | Multi-dimension and strategy-based commodity pricing method and system | |
| CN111784492B (en) | Public opinion analysis and financial early warning method and device, electronic equipment and storage medium | |
| CN115239491A (en) | Futures trading method and computer equipment | |
| Pucheta et al. | Stochastic analysis for short-and long-term forecasting of latin american country risk indexes | |
| CN113327048A (en) | Enterprise portrait calculation method, medium, and program based on big data and multidimensional features | |
| CN114004629A (en) | Data analysis method and device | |
| CN110795537B (en) | Method, device, equipment and medium for determining improvement strategy of target commodity | |
| CN113269610A (en) | Bank product recommendation method and device and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210312 |